Commit graph

1,185,212 commits

Author SHA1 Message Date
Jakub Kicinski
0f19f514de mlx5-updates-2023-02-10
1) From Roi and Mark: MultiPort eswitch support
 
 MultiPort E-Switch builds on newer hardware's capabilities and introduces
 a mode where a single E-Switch is used and all the vports and physical
 ports on the NIC are connected to it.
 
 The new mode will allow in the future a decrease in the memory used by the
 driver and advanced features that aren't possible today.
 
 This represents a big change in the current E-Switch implantation in mlx5.
 Currently, by default, each E-Switch manager manages its E-Switch.
 Steering rules in each E-Switch can only forward traffic to the native
 physical port associated with that E-Switch. While there are ways to target
 non-native physical ports, for example using a bond or via special TC
 rules. None of the ways allows a user to configure the driver
 to operate by default in such a mode nor can the driver decide
 to move to this mode by default as it's user configuration-driven right now.
 
 While MultiPort E-Switch single FDB mode is the preferred mode, older
 generations of ConnectX hardware couldn't support this mode so it was never
 implemented. Now that there is capable hardware present, start the
 transition to having this mode by default.
 
 Introduce a devlink parameter to control MultiPort Eswitch single FDB mode.
 This will allow users to select this mode on their system right now
 and in the future will allow the driver to move to this mode by default.
 
 2) From Jiri: Improvements and fixes for mlx5 netdev's devlink logic
  2.1) Cleanups related to mlx5's devlink port logic
  2.2) Move devlink port registration to be done before netdev alloc
  2.3) Create auxdev devlink instance in the same ns as parent devlink
  2.4) Suspend auxiliary devices only in case of PCI device suspend
 -----BEGIN PGP SIGNATURE-----
 
 iQEzBAABCAAdFiEEGhZs6bAKwk/OTgTpSD+KveBX+j4FAmPsBlsACgkQSD+KveBX
 +j6izAgAtvzimjzOqI8J0FKPW3piRH3VtNWUrlGAf4HHPI0lZMbEV5GIaSI15JeH
 wrIpNXKzaA+6+NZ7TzTwqVO+zyKCQVpRw7xPANIJ2A5dId36jU2AgW06zgzFuYWa
 mXKG8rAUls6E5a5DHTaY/SueWMO9KMOfLzHFXHNsLAKTAw+FoDdlGso5vgvsy44g
 2/N/kDBdzC5n8og934sM7JymthnTVyuGe/Au2w1UUR1MMZ46OQgmhStDwqS5MrzI
 pIi+tWFG9YFM5/J8d13y7X/xMCq26Aw6SZSUhsRfTk0goy0uKaU6yYm8vSbO0mci
 +e0c8bpWL/MzAhMzWj8CwBPeO6knOw==
 =tYc2
 -----END PGP SIGNATURE-----

Merge tag 'mlx5-updates-2023-02-10' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2023-02-10

1) From Roi and Mark: MultiPort eswitch support

MultiPort E-Switch builds on newer hardware's capabilities and introduces
a mode where a single E-Switch is used and all the vports and physical
ports on the NIC are connected to it.

The new mode will allow in the future a decrease in the memory used by the
driver and advanced features that aren't possible today.

This represents a big change in the current E-Switch implantation in mlx5.
Currently, by default, each E-Switch manager manages its E-Switch.
Steering rules in each E-Switch can only forward traffic to the native
physical port associated with that E-Switch. While there are ways to target
non-native physical ports, for example using a bond or via special TC
rules. None of the ways allows a user to configure the driver
to operate by default in such a mode nor can the driver decide
to move to this mode by default as it's user configuration-driven right now.

While MultiPort E-Switch single FDB mode is the preferred mode, older
generations of ConnectX hardware couldn't support this mode so it was never
implemented. Now that there is capable hardware present, start the
transition to having this mode by default.

Introduce a devlink parameter to control MultiPort Eswitch single FDB mode.
This will allow users to select this mode on their system right now
and in the future will allow the driver to move to this mode by default.

2) From Jiri: Improvements and fixes for mlx5 netdev's devlink logic
 2.1) Cleanups related to mlx5's devlink port logic
 2.2) Move devlink port registration to be done before netdev alloc
 2.3) Create auxdev devlink instance in the same ns as parent devlink
 2.4) Suspend auxiliary devices only in case of PCI device suspend

* tag 'mlx5-updates-2023-02-10' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
  net/mlx5: Suspend auxiliary devices only in case of PCI device suspend
  net/mlx5: Remove "recovery" arg from mlx5_load_one() function
  net/mlx5e: Create auxdev devlink instance in the same ns as parent devlink
  net/mlx5e: Move devlink port registration to be done before netdev alloc
  net/mlx5e: Move dl_port to struct mlx5e_dev
  net/mlx5e: Replace usage of mlx5e_devlink_get_dl_port() by netdev->devlink_port
  net/mlx5e: Pass mdev to mlx5e_devlink_port_register()
  net/mlx5: Remove outdated comment
  net/mlx5e: TC, Remove redundant parse_attr argument
  net/mlx5e: Use a simpler comparison for uplink rep
  net/mlx5: Lag, Add single RDMA device in multiport mode
  net/mlx5: Lag, set different uplink vport metadata in multiport eswitch mode
  net/mlx5: E-Switch, rename bond update function to be reused
  net/mlx5e: TC, Add peer flow in mpesw mode
  net/mlx5: Lag, Control MultiPort E-Switch single FDB mode
====================

Link: https://lore.kernel.org/r/20230214221239.159033-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:24:52 -08:00
Deepak R Varma
578292aaec drm/amd/display: Remove duplicate/repeating expression
Remove duplicate or repeating expressions in the if condition
evaluation. Issue identified using doubletest.cocci Coccinelle semantic
patch.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Deepak R Varma <drv@mailo.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:48 -05:00
Arthur Grillo
60b07cf5d3 drm/amd/display: Make variables declaration inside ifdef guard
Make variables declaration inside ifdef guard, as they are only used
inside the same ifdef guard. This remove some of the
-Wunused-but-set-variable warning.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:45 -05:00
Arthur Grillo
01543dcf99 drm/amd/display: Fix excess arguments on kernel-doc
Remove arguments present on kernel-doc that are not present on the
function declaration and add the new ones if present.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:43 -05:00
Arthur Grillo
8dc2507f4b drm/amd/display: Add previously missing includes
Add includes that were previously missing to reduce the number of
-Wmissing-prototypes warnings.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:39 -05:00
Arthur Grillo
877b57c6b5 drm/amd/amdgpu: Add function prototypes to headers
Add function prototypes to headers to reduce the number of
-Wmissing-prototypes warnings.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:36 -05:00
Arthur Grillo
6884b0e407 drm/amd/display: Add function prototypes to headers
Add function prototypes to headers to reduce the number of
-Wmissing-prototypes warnings.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:34 -05:00
Arthur Grillo
1706d9a412 drm/amd/display: Turn global functions into static
Turn global functions that are only used locally into static ones. This
reduces the number of -Wmissing-prototypes warnings.

Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:30 -05:00
Melissa Wen
9a13ff8f52 drm/amd/display: remove unused _calculate_degamma_curve function
We don't use this function anywhere, therefore, remove it.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:26 -05:00
Melissa Wen
62f9286fc2 drm/amd/display: remove unused func declaration from resource headers
The function resource_validate_ctx_update_pointer_after_copy() is
declared in resource.h but never defined, therefore, remove its
declaration from headers.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:23 -05:00
Melissa Wen
af54167d2b drm/amd/display: unset initial value for tf since it's never used
In mod_color_calculate_{degamma/regamma}_params(), a tf variable is
initialized as TRANSFER_FUNCTION_SRGB but tf is only used after tf =
input->tf, therefore, better to just remove this initial value and avoid
misleading interpretations.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:20 -05:00
Melissa Wen
6ff387e889 drm/amd/display: camel case cleanup in color_gamma file
Rename mapUserRamp to map_user_ramp and doClamping to do_clamping

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:15 -05:00
Lijo Lazar
18c4e319db drm/amd/pm: Allocate dummy table only if needed
Only Navi1x requires dummy read workaround. Allocate the table in VRAM
only for Navi1x.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:24:08 -05:00
Qingqing Zhuo
d8f3b3951f Revert "drm/amd/display: enable DPG when disabling plane for phantom pipe"
This reverts commit d47d2f9392.

regression detected by the change. Revert until
fix is available.

Signed-off-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:23:13 -05:00
Aric Cyr
2f5d28ae2d drm/amd/display: Promote DAL to 3.2.223
This version brings along the following:
- Move domain power control to DMCUB for DCN314
- Enable P-state validation check for DCN314
- Add support for multiple overlay planes
- Fixes in prefetch, k1 k2 divider programming and more
- Code cleanup

Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:23:06 -05:00
Ayush Gupta
d2994b25e0 drm/amd/display: temporary fix for page faulting
This reverts a part of the
commit 826e7ffaf0
 ("drm/amd/display: [FW Promotion] Release 0.0.153.0")

Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Ayush Gupta<ayugupta@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:21:07 -05:00
Jakub Kicinski
dee4bf7167 Merge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:

====================
Intel Wired LAN Driver Updates 2023-02-14 (ixgbe, i40e)

This series contains updates to ixgbe and i40e drivers.

Jason Xing corrects comparison of frame sizes for setting MTU with XDP on
ixgbe and adjusts frame size to account for a second VLAN header on ixgbe
and i40e.

* '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
  ixgbe: add double of VLAN header when computing the max MTU
  i40e: add double of VLAN header when computing the max MTU
  ixgbe: allow to increase MTU to 3K with XDP enabled
====================

Link: https://lore.kernel.org/r/20230214185146.1305819-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:20:58 -08:00
Mustapha Ghaddar
8e5cfe547b drm/amd/display: upstream link_dp_dpia_bw.c
[WHY & HOW]
- make link_dp_dpia_bw.c available for linux.
- add the verify link peak bw
- clean up code and comment format.

Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Mustapha Ghaddar <mghaddar@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:20:49 -05:00
Nicholas Kazlauskas
b7c67f7240 drm/amd/display: Disable HUBP/DPP PG on DCN314 for now
[Why]
The DMCUB implementation required to workaround corruption is
not currently stable and may cause intermittent corruption or hangs.

[How]
Disable PG until the sequence is stable.

Reviewed-by: Hansen Dsouza <hansen.dsouza@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:20:38 -05:00
Melissa Wen
94369589e4 drm/amd/display: clean code-style issues in dcn30_set_mpc_shaper_3dlut
This function has many conditions and all code style issues (identation,
missing braces, etc.) make reading it really annoying.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:20:32 -05:00
Melissa Wen
f52231c4a8 drm/amd/display: ident braces in dcn30_acquire_post_bldn_3dlut correctly
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:20:27 -05:00
Jack Xiao
dc907c9db8 drm/amd/amdgpu: fix warning during suspend
Freeing memory was warned during suspend.
Move the self test out of suspend.

Link: https://bugzilla.redhat.com/show_bug.cgi?id=2151825
Cc: jfalempe@redhat.com
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-and-tested-by: Evan Quan <evan.quan@amd.com>
Tested-by: Jocelyn Falempe <jfalempe@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2023-02-15 22:19:43 -05:00
Jakub Kicinski
388a9c907a Merge branch 'devlink-cleanups-and-move-devlink-health-functionality-to-separate-file'
Moshe Shemesh says:

====================
devlink: cleanups and move devlink health functionality to separate file

This patchset moves devlink health callbacks, helpers and related code
from leftover.c to new file health.c. About 1.3K LoC are moved by this
patchset, covering all devlink health functionality.

In addition this patchset includes a couple of small cleanups in devlink
health code and documentation update.
====================

Link: https://lore.kernel.org/r/1676392686-405892-1-git-send-email-moshe@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:46 -08:00
Moshe Shemesh
d0ab772c1f devlink: Fix TP_STRUCT_entry in trace of devlink health report
Fix a bug in trace point definition for devlink health report, as
TP_STRUCT_entry of reporter_name should get reporter_name and not msg.

Note no fixes tag as this is a harmless bug as both reporter_name and
msg are strings and TP_fast_assign for this entry is correct.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
c745cfb27a devlink: Update devlink health documentation
Update devlink-health.rst file:
- Add devlink formatted message (fmsg) API documentation.
- Add auto-dump as a condition to do dump once error reported.
- Expand OOB to clarify this acronym.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
12af29e779 devlink: Move health common function to health file
Now that all devlink health callbacks and related code are in file
health.c move common health functions and devlink_health_reporter struct
to be local in health.c file.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
c9311ee13f devlink: Move devlink health test to health file
Move devlink health report test callback from leftover.c to health.c. No
functional change in this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
7004c6c457 devlink: Move devlink health dump to health file
Move devlink health report dump callbacks and related code from
leftover.c to health.c. No functional change in this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
a929df7fd9 devlink: Move devlink fmsg and health diagnose to health file
Devlink fmsg (formatted message) is used by devlink health diagnose,
dump and drivers which support these devlink health callbacks.
Therefore, move devlink fmsg helpers and related code to file health.c.
Move devlink health diagnose to file health.c. No functional change in
this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
55b9b24968 devlink: Move devlink health report and recover to health file
Move devlink health report helper and recover callback and related code
from leftover.c to health.c. No functional change in this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
db6b5f3ec4 devlink: Move devlink health get and set code to health file
Move devlink health get and set callbacks and related code from
leftover.c to health.c. No functional change in this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
bfd4e6a5db devlink: health: Fix nla_nest_end in error flow
devlink_nl_health_reporter_fill() error flow calls nla_nest_end(). Fix
it to call nla_nest_cancel() instead.

Note the bug is harmless as genlmsg_cancel() cancel the entire message,
so no fixes tag added.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Moshe Shemesh
b4740e3a81 devlink: Split out health reporter create code
Move devlink health reporter create/destroy and related dev code to new
file health.c. This file shall include all callbacks and functionality
that are related to devlink health.

In addition, fix kdoc indentation and make reporter create/destroy kdoc
more clear. No functional change in this patch.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:15:44 -08:00
Lorenzo Bianconi
b6a4103c35 ice: update xdp_features with xdp multi-buff
Now ice driver supports xdp multi-buffer so add it to xdp_features.
Check vsi type before setting xdp_features flag.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/r/8a4781511ab6e3cd280e944eef69158954f1a15f.1676385351.git.lorenzo@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:10:32 -08:00
Lorenzo Bianconi
9dd6e53ef6 i40e: check vsi type before setting xdp_features flag
Set xdp_features flag just for I40E_VSI_MAIN vsi type since XDP is
supported just in this configuration.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Link: https://lore.kernel.org/r/f2b537f86b34fc176fbc6b3d249b46a20a87a2f3.1676405131.git.lorenzo@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-02-15 19:07:54 -08:00
Alexander Lobakin
6c20822fad bpf, test_run: fix &xdp_frame misplacement for LIVE_FRAMES
&xdp_buff and &xdp_frame are bound in a way that

xdp_buff->data_hard_start == xdp_frame

It's always the case and e.g. xdp_convert_buff_to_frame() relies on
this.
IOW, the following:

	for (u32 i = 0; i < 0xdead; i++) {
		xdpf = xdp_convert_buff_to_frame(&xdp);
		xdp_convert_frame_to_buff(xdpf, &xdp);
	}

shouldn't ever modify @xdpf's contents or the pointer itself.
However, "live packet" code wrongly treats &xdp_frame as part of its
context placed *before* the data_hard_start. With such flow,
data_hard_start is sizeof(*xdpf) off to the right and no longer points
to the XDP frame.

Instead of replacing `sizeof(ctx)` with `offsetof(ctx, xdpf)` in several
places and praying that there are no more miscalcs left somewhere in the
code, unionize ::frm with ::data in a flex array, so that both starts
pointing to the actual data_hard_start and the XDP frame actually starts
being a part of it, i.e. a part of the headroom, not the context.
A nice side effect is that the maximum frame size for this mode gets
increased by 40 bytes, as xdp_buff::frame_sz includes everything from
data_hard_start (-> includes xdpf already) to the end of XDP/skb shared
info.
Also update %MAX_PKT_SIZE accordingly in the selftests code. Leave it
hardcoded for 64 bit && 4k pages, it can be made more flexible later on.

Minor: align `&head->data` with how `head->frm` is assigned for
consistency.
Minor #2: rename 'frm' to 'frame' in &xdp_page_head while at it for
clarity.

(was found while testing XDP traffic generator on ice, which calls
 xdp_convert_frame_to_buff() for each XDP frame)

Fixes: b530e9e106 ("bpf: Add "live packet" mode for XDP in BPF_PROG_RUN")
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Link: https://lore.kernel.org/r/20230215185440.4126672-1-aleksander.lobakin@intel.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
2023-02-15 17:39:36 -08:00
Jason Gunthorpe
b4ff830eca iommufd: Do not add the same hwpt to the ioas->hwpt_list twice
The hwpt is added to the hwpt_list only during its creation, it is never
added again. This hunk is some missed leftover from rework. Adding it
twice will corrupt the linked list in some cases.

It effects HWPT specific attachment, which is something the test suite
cannot cover until we can create a legitimate struct device with a
non-system iommu "driver" (ie we need the bus removed from the iommu code)

Cc: stable@vger.kernel.org
Fixes: e8d5721003 ("iommufd: Add kAPI toward external drivers for physical devices")
Link: https://lore.kernel.org/r/1-v1-4336b5cb2fe4+1d7-iommufd_hwpt_jgg@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reported-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2023-02-15 21:37:48 -04:00
Steven Rostedt (Google)
8261ef2eb3 tracing: Add BUILD_BUG() to make sure stacktrace fits in strings
The max string length for a histogram variable is 256 bytes. The max depth
of a stacktrace is 16. With 8byte words, that's 16 * 8 = 128. Which can
easily fit in the string variable. The histogram stacktrace is being
stored in the string value (with the given max length), with the
assumption it will fit. To make sure that this is always the case (in the
case that the stack trace depth increases), add a BUILD_BUG_ON() to test
this.

Link: https://lore.kernel.org/linux-trace-kernel/20230214002418.0103b9e765d3e5c374d2aa7d@kernel.org/

Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
2023-02-15 20:25:38 -05:00
Tom Zanussi
fc1a9dc101 tracing/histogram: Don't use strlen to find length of stacktrace variables
Because stacktraces are saved in dynamic strings,
trace_event_raw_event_synth() uses strlen to determine the length of
the stack.  Stacktraces may contain 0-bytes, though, in the saved
addresses, so the length found and passed to reserve() will be too
small.

Fix this by using the first unsigned long in the stack variables to
store the actual number of elements in the stack and have
trace_event_raw_event_synth() use that to determine the length of the
stack.

Link: https://lkml.kernel.org/r/1ed6906cd9d6477ef2bd8e63c61de20a9ffe64d7.1676063532.git.zanussi@kernel.org

Signed-off-by: Tom Zanussi <zanussi@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
2023-02-15 19:59:09 -05:00
Andrii Nakryiko
d964f09af4 Merge branch 'New benchmark for hashmap lookups'
Anton Protopopov says:

====================

Add a new benchmark for hashmap lookups and fix several typos.

In commit 3 I've patched the bench utility so that now command line options
can be reused by different benchmarks.

The benchmark itself is added in the last commit 7. I was using this benchmark
to test map lookup productivity when using a different hash function [1]. When
run with --quiet, the results can be easily plotted [2].  The results provided
by the benchmark look reasonable and match the results of my different
benchmarks (requiring to patch kernel to get actual statistics on map lookups).

Links:
  [1] https://fosdem.org/2023/schedule/event/bpf_hashing/
  [2] https://github.com/aspsk/bpf-bench/tree/master/hashmap-bench

Changes,
v1->v2:
- percpu_times_index[] is of wrong size (Martin)
- use base 0 for strtol (Andrii)
- just use -q without argument (Andrii)
- use less hacks when parsing arguments (Andrii)
====================

Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
2023-02-15 16:29:32 -08:00
Anton Protopopov
f371f2dc53 selftest/bpf/benchs: Add benchmark for hashmap lookups
Add a new benchmark which measures hashmap lookup operations speed.  A user can
control the following parameters of the benchmark:

    * key_size (max 1024): the key size to use
    * max_entries: the hashmap max entries
    * nr_entries: the number of entries to insert/lookup
    * nr_loops: the number of loops for the benchmark
    * map_flags The hashmap flags passed to BPF_MAP_CREATE

The BPF program performing the benchmarks calls two nested bpf_loop:

    bpf_loop(nr_loops/nr_entries)
            bpf_loop(nr_entries)
                     bpf_map_lookup()

So the nr_loops determines the number of actual map lookups. All lookups are
successful.

Example (the output is generated on a AMD Ryzen 9 3950X machine):

    for nr_entries in `seq 4096 4096 65536`; do echo -n "$((nr_entries*100/65536))% full: "; sudo ./bench -d2 -a bpf-hashmap-lookup --key_size=4 --nr_entries=$nr_entries --max_entries=65536 --nr_loops=1000000 --map_flags=0x40 | grep cpu; done
    6% full: cpu01: lookup 50.739M ± 0.018M events/sec (approximated from 32 samples of ~19ms)
    12% full: cpu01: lookup 47.751M ± 0.015M events/sec (approximated from 32 samples of ~20ms)
    18% full: cpu01: lookup 45.153M ± 0.013M events/sec (approximated from 32 samples of ~22ms)
    25% full: cpu01: lookup 43.826M ± 0.014M events/sec (approximated from 32 samples of ~22ms)
    31% full: cpu01: lookup 41.971M ± 0.012M events/sec (approximated from 32 samples of ~23ms)
    37% full: cpu01: lookup 41.034M ± 0.015M events/sec (approximated from 32 samples of ~24ms)
    43% full: cpu01: lookup 39.946M ± 0.012M events/sec (approximated from 32 samples of ~25ms)
    50% full: cpu01: lookup 38.256M ± 0.014M events/sec (approximated from 32 samples of ~26ms)
    56% full: cpu01: lookup 36.580M ± 0.018M events/sec (approximated from 32 samples of ~27ms)
    62% full: cpu01: lookup 36.252M ± 0.012M events/sec (approximated from 32 samples of ~27ms)
    68% full: cpu01: lookup 35.200M ± 0.012M events/sec (approximated from 32 samples of ~28ms)
    75% full: cpu01: lookup 34.061M ± 0.009M events/sec (approximated from 32 samples of ~29ms)
    81% full: cpu01: lookup 34.374M ± 0.010M events/sec (approximated from 32 samples of ~29ms)
    87% full: cpu01: lookup 33.244M ± 0.011M events/sec (approximated from 32 samples of ~30ms)
    93% full: cpu01: lookup 32.182M ± 0.013M events/sec (approximated from 32 samples of ~31ms)
    100% full: cpu01: lookup 31.497M ± 0.016M events/sec (approximated from 32 samples of ~31ms)

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-8-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
a237dda05e selftest/bpf/benchs: Print less if the quiet option is set
The bench utility will print

    Setting up benchmark '<bench-name>'...
    Benchmark '<bench-name>' started.

on startup to stdout. Suppress this output if --quiet option if given. This
makes it simpler to parse benchmark output by a script.

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-7-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
90c22503cd selftest/bpf/benchs: Make quiet option common
The "local-storage-tasks-trace" benchmark has a `--quiet` option. Move it to
the list of common options, so that the main code and other benchmarks can use
(new) env.quiet variable. Patch the run_bench_local_storage_rcu_tasks_trace.sh
helper script accordingly.

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-6-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
9644546260 selftest/bpf/benchs: Remove an unused header
The benchs/bench_bpf_hashmap_full_update.c doesn't set a custom argp,
so it shouldn't include the <argp.h> header.

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-5-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
22ff7aeaa9 selftest/bpf/benchs: Enhance argp parsing
To parse command line the bench utility uses the argp_parse() function. This
function takes as an argument a parent 'struct argp' structure which defines
common command line options and an array of children 'struct argp' structures
which defines additional command line options for particular benchmarks. This
implementation doesn't allow benchmarks to share option names, e.g., if two
benchmarks want to use, say, the --option option, then only one of them will
succeed (the first one encountered in the array).  This will be convenient if
same option names could be used in different benchmarks (with the same
semantics, e.g., --nr_loops=N).

Fix this by calling the argp_parse() function twice. The first call is the same
as it was before, with all children argps, and helps to find the benchmark name
and to print a combined help message if anything is wrong.  Given the name, we
can call the argp_parse the second time, but now the children array points only
to a correct benchmark thus always calling the correct parsers. (If there's no
a specific list of arguments, then only one call to argp_parse will be done.)

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-4-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
2f1c59637f selftest/bpf/benchs: Make a function static in bpf_hashmap_full_update
The hashmap_report_final callback function defined in the
benchs/bench_bpf_hashmap_full_update.c file should be static.

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-3-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Anton Protopopov
4db98ab445 selftest/bpf/benchs: Fix a typo in bpf_hashmap_full_update
To call the bpf_hashmap_full_update benchmark, one should say:

    bench bpf-hashmap-ful-update

The patch adds a missing 'l' to the benchmark name.

Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230213091519.1202813-2-aspsk@isovalent.com
2023-02-15 16:29:31 -08:00
Alexei Starovoitov
3538a0fbbd Merge branch 'Use __GFP_ZERO in bpf memory allocator'
Hou Tao says:

====================

From: Hou Tao <houtao1@huawei.com>

Hi,

The patchset tries to fix the hard-up problem found when checking how htab
handles element reuse in bpf memory allocator. The immediate reuse of
freed elements will reinitialize special fields (e.g., bpf_spin_lock) in
htab map value and it may corrupt lookup procedure with BFP_F_LOCK flag
which acquires bpf-spin-lock during value copying, and lead to hard-lock
as shown in patch #2. Patch #1 fixes it by using __GFP_ZERO when allocating
the object from slab and the behavior is similar with the preallocated
hash-table case. Please see individual patches for more details. And comments
are always welcome.

Regards,

Change Log:
v1:
  * Use __GFP_ZERO instead of ctor to avoid retpoline overhead (from Alexei)
  * Add comments for check_and_init_map_value() (from Alexei)
  * split __GFP_ZERO patches out of the original patchset to unblock
    the development work of others.

RFC: https://lore.kernel.org/bpf/20221230041151.1231169-1-houtao@huaweicloud.com
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2023-02-15 15:40:06 -08:00
Hou Tao
f88da2d46c selftests/bpf: Add test case for element reuse in htab map
The reinitialization of spin-lock in map value after immediate reuse may
corrupt lookup with BPF_F_LOCK flag and result in hard lock-up, so add
one test case to demonstrate the problem.

Signed-off-by: Hou Tao <houtao1@huawei.com>
Link: https://lore.kernel.org/r/20230215082132.3856544-3-houtao@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2023-02-15 15:40:06 -08:00
Hou Tao
997849c4b9 bpf: Zeroing allocated object from slab in bpf memory allocator
Currently the freed element in bpf memory allocator may be immediately
reused, for htab map the reuse will reinitialize special fields in map
value (e.g., bpf_spin_lock), but lookup procedure may still access
these special fields, and it may lead to hard-lockup as shown below:

 NMI backtrace for cpu 16
 CPU: 16 PID: 2574 Comm: htab.bin Tainted: G             L     6.1.0+ #1
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
 RIP: 0010:queued_spin_lock_slowpath+0x283/0x2c0
 ......
 Call Trace:
  <TASK>
  copy_map_value_locked+0xb7/0x170
  bpf_map_copy_value+0x113/0x3c0
  __sys_bpf+0x1c67/0x2780
  __x64_sys_bpf+0x1c/0x20
  do_syscall_64+0x30/0x60
  entry_SYSCALL_64_after_hwframe+0x46/0xb0
 ......
  </TASK>

For htab map, just like the preallocated case, these is no need to
initialize these special fields in map value again once these fields
have been initialized. For preallocated htab map, these fields are
initialized through __GFP_ZERO in bpf_map_area_alloc(), so do the
similar thing for non-preallocated htab in bpf memory allocator. And
there is no need to use __GFP_ZERO for per-cpu bpf memory allocator,
because __alloc_percpu_gfp() does it implicitly.

Fixes: 0fd7c5d433 ("bpf: Optimize call_rcu in non-preallocated hash map.")
Signed-off-by: Hou Tao <houtao1@huawei.com>
Link: https://lore.kernel.org/r/20230215082132.3856544-2-houtao@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2023-02-15 15:40:06 -08:00